Remote stereocontrol in the Nazarov reaction: A new approach to the core of roseophilin

Article abstract of DOI:10.1021/jo0504058

Three different procedures are compared to obtain properly substituted divinyl ketones in which one of the double bonds is embedded in a five-membered heterocyclic structure and therefore suitable to produce cyclopenta-fused pyrrole derivatives by the acid-catalyzed Nazarov reaction. These, on treatment with TFA, afforded 2,4-cis-disubstituted 2,3,4,5-tetrahydro-1H-cyclopenta[e] pyrrol-6-ones with high stereo-control. One of these Nazarov products was oxidized to the corresponding 4,5-dihydro-1H-cyclopenta[e]pyrrol-6-one derivative, thus obtaining an enantiopure key intermediate in the total synthesis of roseophilin.

Full text of DOI:10.1021/jo0504058

SCHEME 1
Remote Stereocontrol in the Nazarov
Reaction: A New Approach to the Core of
Roseophilin
Ernesto G. Occhiato,*,† Cristina Prandi,*,‡
Alessandro Ferrali,† and Antonio Guarna†
Dipartimento di Chimica Organica “U. Schiff”, Polo
Scientifico - Universita` di Firenze, Via della Lastruccia 13,
I-50019 Sesto Fiorentino, Italy, and Dipartimento di
Chimica Generale ed Organica Applicata, Universita` di
Torino, Corso Massimo D’Azeglio, 48, I-10125 Torino, Italy
ernesto.occhiato@unifi.it; cristina.prandi@unito.it
Received March 3, 2005
pound, can be prepared in an earlier phase which is
followed by the RCM of the terminal double bonds
present on the two chains at C2 and C5.2a,b,g,k This
sequence has been first used by Fuchs in his synthesis
of 2, which required, however, 13 steps to prepare key
intermediate 5 in racemic form.2b The stimulus to revisit
Fuchs’ formal synthesis of roseophilin came from our
recent finding that the Nazarov reaction of divinyl
ketones in which one of the double bonds is embedded
in a properly substituted N-heterocyclic structure pro-
ceeds in highly stereocontrolled fashion to give cis-
disubstituted cyclopenta-fused heterocycles.4 On these
grounds, we envisioned a faster route to obtain bicyclic
ketopyrrole 5 in enantiopure form by electrocyclization
of pyrroline 7 (Scheme 1) in which the correctly oriented
buten-3-yl chain on the heterocycle serves the purpose
of controlling the absolute stereochemistry of the C4 atom
bearing the isopropyl group in the Nazarov product 6
while being already set for the RCM following Fuchs’
strategy from 5.
Three different procedures are compared to obtain properly
substituted divinyl ketones in which one of the double bonds
is embedded in a five-membered heterocyclic structure and
therefore suitable to produce cyclopenta-fused pyrrole de-
rivatives by the acid-catalyzed Nazarov reaction. These, on
treatment with TFA, afforded 2,4-cis-disubstituted 2,3,4,5-
tetrahydro-1H-cyclopenta[b]pyrrol-6-ones with high stereo-
control. One of these Nazarov products was oxidized to the
corresponding 4,5-dihydro-1H-cyclopenta[b]pyrrol-6-one de-
rivative, thus obtaining an enantiopure key intermediate in
the total synthesis of roseophilin.
Roseophilin (1, Scheme 1) is a macrocyclic pigment
isolated from Streptomyces griseoviridis that exhibits
potent cytotoxicity against human cancer cell lines.1 This,
and the unique ansa-bridged cyclopenta[b]pyrrole struc-
tural core of roseophilin, have stimulated the interest of
several groups in the partial or total synthesis of this
heterocycle.2 As pointed out by Fu¨rstner in his exhaustive
review on the chemistry and biology of roseophilin,3 all
synthetic approaches toward 1 are based on the same
disconnection that includes the condensation of the
macrotricyclic cyclopenta-fused pyrrole 2 with the
heterocyclic portion 3. The procedures by which inter-
mediate 2 has been synthesized can in turn be grouped
in two general strategies. The most employed relies
on the construction of the ansa bridge moiety which
anchors suitable functionalities to eventually assem-
bly the 1-azafulvene moiety by diverse cyclization
reactions.2c-f,h-j,l-o Alternatively, the isopropyl-substitut-
ed ketopyrrole 4 (Scheme 1), or a closely related com-
(2) (a) Kim, S. H.; Fuchs, P. L. Tetrahedron Lett. 1996, 37, 2545-
2548. (b) Kim, S. H.; Figueroa, I.; Fuchs, P. L. Tetrahedron Lett. 1997,
38, 2601-2604. (c) Fu¨rstner, A.; Weintritt, H. J. Am. Chem. Soc. 1997,
119, 2944-2945. (d) Fu¨rstner, A.; Weintritt, H. J. Am. Chem. Soc.
1998, 120, 2817-2825. (e) Mochizuki, T.; Itoh, E.; Shibata, N.;
Nakatami, S.; Katoh, T.; Terashima, S. Tetrahedron Lett. 1998, 39,
6911-6914. (f) Robertson, J.; Hatley, R. J. D. Chem. Commun. 1999,
1455-1456. (g) Fagan, M. A.; Knight, D. W. Tetrahedron Lett. 1999,
40, 6117-6120. (h) Fu¨rstner, A.; Gastner, T.; Weintritt, H. J. Org.
Chem. 1999, 64, 2361-2366. (i) Harrington, P. E.; Tius, M. A. Org.
Lett. 1999, 1, 649-651. (j) Robertson, J.; Hatley, R. J. D.; Watkin, D.
J. J. Chem. Soc., Perkin Trans. 1 2000, 3389-3396. (k) Bamford, S.
J.; Luker, T.; Speckamp, W. N.; Hiemstra, H. Org. Lett. 2000, 2, 1157-
1160. (l) Trost, B. M.; Doherty, G. A. J. Am. Chem. Soc. 2000, 122,
3801-3810. (m) Harrington, P. E.; Tius, M. A. J. Am. Chem. Soc.. 2001,
123, 8509-8514. (n) Boger, D. L.; Hong, J. J. Am. Chem. Soc. 2001,
123, 8515-8519. (o) Viseux, E. M. E.; Parsons, P. J.; Pavey, J. B. J.
Carter, C. M. Pinto, I. Synlett 2003, 1856-1858.
* To whom correspondence should be addressed. (E.G.O.) Tel: +39-
055-4573480. Fax: +39-055-4573531. (C.P.) Tel: +39-011-6707641.
Fax: 0116707642.
(3) Fu¨rstner, A. Angew. Chem., Int. Ed. 2003, 42, 3582-3603.
(4) (a) Occhiato, E. G.; Prandi, C.; Ferrali, A.; Guarna, A.; Venturello,
P. J. Org. Chem. 2003, 68, 9728-9741. (b) Prandi, C.; Ferrali, A.;
Guarna, A.; Venturello, P.; Occhiato, E. G. J. Org. Chem. 2004, 69,
7705-7709.
† Universita` di Firenze.
‡ Universita` di Torino.
(1) Hayakawa, Y.; Kawakami, K.; Seto, H.; Furihata, K. Tetrahedron
Lett. 1992, 33, 2701-2704.
10.1021/jo0504058 CCC: $30.25 © 2005 American Chemical Society
Published on Web 04/19/2005
4542
J. Org. Chem. 2005, 70, 4542-4545

SCHEME 2a
SCHEME 4a
a Key: (a) TsCl, Et3N, DMAP, CH2Cl2, 25 °C, 16 h; (b) allyl-
magnesium bromide, THF, 0 f 25 °C, 6 h; (c) TsCl, LHMDS, THF,
-20 °C, 1 h; (d) PhNTf2, KHMDS, THF, -78 °C, 1 h; then to 0 °C;
(e) 13, 4% (Ph3P)2PdCl2, 2 M Na2CO3, THF, 55 °C, 4 h.
SCHEME 3a
a Key: (a) PhNTf2, KHMDS, THF, -78 °C, 1 h; then to 0 °C;
(b) 10% Pd(OAc)2, 20% Ph3P, CO, DMF 10 min; then Et3N, MeOH,
40 °C, 4 h; (c) CH3PO(OCH3)2, n-BuLi, THF, -78 °C, 30 min; then
19-21, 25 °C, 4 h; (d) 2-methylpropanal, DIPEA, LiCl, MeCN, 25
°C, 24 h; (e) TFA, 25 °C, 20 h; (f) DDQ, toluene, reflux, 5 h.
a Key: (a) triethyl orthoformate, NH4NO3, EtOH; (b) t-BuOK,
n-BuLi, THF, -95 °C, 2 h; (c) triisopropylborate, 25 °C; (d) pinacol,
toluene, 25 °C.
Following our earlier studies,4a,b we initially relied, for
the preparation of 7, on the hydrolysis of ethoxytriene
14 (Scheme 2) obtained by Pd-catalyzed coupling of vinyl
triflate 12 with R-ethoxydienylboronate 13, this in turn
prepared from the corresponding aldehyde 15 as depicted
in Scheme 3.5
acidic conditions (Amberlyst 15 resin in dichloromethane
or 0.02 M HCl in MeOH-H2O) as in other cases.4,10c
Later, we found that the hydrolysis of this robust
ethoxydiene moiety requires harsh conditions (heating
at 80 °C with 20% H2SO4) to occur,11 not compatible with
the presence of both the delicate heterocycle moiety of
14 and the terminal olefin, and this convergent approach
to 7 was abandoned.
Two alternative routes were envisioned to prepare 7.
Initially, we started a linear sequence (Scheme 4) from
enantiopure (R)-pyrrolidinone 17swith R ) Me, to assess
the proceduresand enantiopure O-TBDMS-protected (S)-
hydroxymethyl pyrrolidinone 18.
The latter substrate was chosen in anticipation of the
need to incorporate the buten-3-yl chain in a later stage
of the synthesis. After transformation into the known
vinyl triflates,4b both substrates were carbonylated (10%
Pd(OAc)2, Ph3P, Et3N, CO)12 in the presence of MeOH to
give 19 and 20 in 63% and 53% yield, respectively.
Different carbonylation conditions13 provided the methyl
esters in comparable yields. We tried also the 2-methyl
pyrrolidinone-derived vinyl phosphate,12 instead of the
triflate, as the substrate for the carbonylation, but it
underwent hydrolysis during the reaction giving back the
Reaction of allylmagnesium bromide (5 equiv) with
enantiopure (S)-9,6 under refluxing conditions in THF,7
furnished 10 (61% yield after chromatography) which was
N tosylated (n-BuLi, TsCl) to give 11 (59%). Attempts at
improving the yield of allylation by carrying out the
reaction in the presence of Cu(I) salts (CuI, CuCN) under
various conditions8 were not successful. A better yield of
11 was obtained when the allylation on crude 9 was
carried out at room temperature (6 h) with 2 equiv of
the Grignard reagent and the next tosylation performed
directly on crude 10 by using LHMDS as a base (56%
over three steps). After N-tosylation, lactam 11 was
converted into vinyl triflate 12 which was directly coupled
with 13 without prior purification.9 The Pd-catalyzed
coupling4,10 proceeded smoothly providing 14 in 45% yield
(over two steps) after chromatography. Unfortunately,
the hydrolysis of the ethoxydiene moiety of 14 to furnish
the corresponding vinyl ketone 7 did not occur under mild
(5) Balma Tivola, P.; Deagostino, A.; Prandi, C.; Venturello, P. Org.
Lett. 2002, 4, 1275-1277.
(6) Nakagawa, Y.; Matsumoto, K.; Tomioka, K. Tetrahedron 2000,
56, 2857-2864.
(10) (a) Occhiato, E. G.; Trabocchi, A.; Guarna, A. Org. Lett. 2000,
2, 1241-1242. (b) Occhiato, E. G.; Trabocchi, A.; Guarna, A. J. Org.
Chem. 2001, 66, 2459-2465. (c) Occhiato, E. G.; Prandi, C.; Ferrali,
A.; Guarna, A.; Deagostino, A.; Venturello, P. J. Org. Chem. 2002, 67,
7144-7146.
(7) Taber, D. F.; Deker, P. B.; Gaul, M. D. J. Am. Chem. Soc. 1987,
109, 7488-7494.
(8) (a) 43% yield in the presence of CuI. Hori, K.; Hikage, N.;
Inagaki, A.; Mori, S.; Keiichi, N.; Yoshii, E. J. Org. Chem. 1992, 57,
2888-2902. (b) 19% yield in the presence of CuCN. Yoda, H.; Katoh,
H.; Ujihara, Y.; Takabe, K. Tetrahedron Lett. 2001, 42, 2509-2512.
(9) Five-membered lactam-derived vinyl triflates are prone to rapid
decomposition on standing. For a review on the chemistry of lactam-
derived vinyl triflates, see: Occhiato, E. G. Mini-Rev. Org. Chem. 2004,
1, 149-162.
(11) Quintard, J.-P.; Elissondo, B.; Hattich, T.; Pereyre, M. J.
Organomet. Chem. 1985, 285, 149-162.
(12) Nicolaou, K. C.; Shi, G.-Q.; Namoto, K.; Bernal, F. Chem.
Commun. 1998, 1757-1758.
(13) (a) Toyooka, N.; Okumura, M.; Takahata, H.; Nemoto, H.
Tetrahedron 1999, 55, 10673-10684. (b) Luker, T.; Hiemstra, H.;
Speckamp, W. N. Tetrahedron Lett. 1996, 37, 8257-8260.
J. Org. Chem, Vol. 70, No. 11, 2005 4543

SCHEME 5
FIGURE 1. Diagnostic NOE relationships in compounds 27,
28, and 6.
SCHEME 6a
lactam. These esters were then transformed into the
corresponding Horner-Emmons-Wadsworth reagents
22 (43%) and 23 (37%) according to the procedure
described by Chiu for related carbacyclic systems14 but,
despite several efforts to improve the reaction conditions,
never with the same excellent yields reported for those.
Olefination of isobutyric aldehyde with 22 and 23 pro-
ceeded smoothly to give the required dienones 25 and
26 which were directly used without purification in the
next step. The coupling constant values of the olefinic
protons (15.4 Hz in 25 and 16.1 Hz in 26) are in
accordance with the expected trans stereochemistry of
the newly formed double bond. The electrocyclization
reaction of these divinyl ketones was carried out by
dissolving the substrates in cold TFA and then allowing
the mixture to sit at room temperature.4 The reactions
were complete in 20 h and furnished cyclopenta-fused
systems 27 and 28 in 40% and 27% yield, respectively,
after chromatography. Similar results in terms of yields
were obtained by carrying out the reaction on 25 in a
0.2 M CF3SO3H solution in CH2Cl2. We were pleased to
find that the methyl-substituted ketopyrroline 27 was
obtained, albeit in low yield, as a single cis diastereomer
(stereochemical assignment based on NOESY 1D and 2D
studies, Figure 1),15 whereas 28 was obtained as a 7:1
mixture of inseparable cis (major) and trans (minor)
diastereomers.15 These results are in accordance with
those previously reported: the reaction proceeds via a
clockwise conrotation which involves the less hindered
face of the endocyclic double bond, i.e., that opposite to
the axially oriented 2-alkyl group on the heterocycle.4b
The substituent on the external double bond (in this case
the isopropyl group) seems to have only a small influence
on the choice of the sense of conrotation in the Nazarov
reaction.
a Key: (a) PhNTf2, KHMDS, THF, -78 °C, 1 h; then to 0 °C;
(b) Me6Sn2, 5% Pd(MeCN)2Cl2, Ph3As, THF, 40 °C, 4 h; (c) 5%
(Ph3P)4Pd, toluene, reflux.
cation 29, formed during the electrocyclization reaction,
thus leading to an undesired tricyclic product 30 (Scheme
5).16
Compound 7 was prepared from enantiopure pyrroli-
dinone 11 by applying the above sequence (Scheme 4),
but once again, low yields (41%) were obtained in the
formation of the corresponding phosphonate 24. There-
fore we experimented with a more convergent procedure
for the preparation of 7 as depicted in Scheme 6. We first
converted triflate 12 into the corresponding stannane 31
by Pd-catalyzed coupling reaction with Me6Sn2,17 and
then crude stannane 31 was coupled with 4-methyl-2-
pentenoyl chloride (32) in the presence of (Ph3P)4Pd as a
catalyst in refluxing toluene. To our knowledge, the Stille
coupling of lactam-derived vinyl stannanes with acyl
chlorides has never been reported so far, and in the
present case it provided pure 7 in 24% over three steps.
To complete the synthesis, compound 7 was treated
with pure TFA and the electrocyclization was complete
1
in 20 h (by TLC). It was a relief to find, by H NMR
analysis, that the crude reaction mixture contained
compound 6 as the major component in an approximately
4:1 ratio with a second compound. The presence in the
spectrum of olefinic signals at 5.62, 5.50, and 5.34 ppm
could actually be accounted for by the formation of 30 as
the minor product; however, we were unable to isolate
this byproduct in order to unequivocally assign the
structure. As in the case of 2-methyl-substituted com-
pound 27, (2R,4S)-6 was also obtained as a single
diastereomer (43% yield after chromatography) whose
2,4-cis relative stereochemistry was established by NOE-
SY studies (Figure 1).15 Compound 6 was finally subjected
to oxidation by DDQ18 yielding our target compound
(4S)-5 in 48% after chromatography.19 In total, this key
The low yield and the formation of a mixture of
diastereomers in the cyclization of 2-hydroxymethyl-
substituted dienone 26 persuaded us to insert the req-
uisite homoallyl chain in an earlier stage of the synthesis
of 5. Although this could be an apparently obvious
strategy, a potential problem arises from this choice: the
terminal double bond could trap the [3.3.0] bicylic oxyallyl
(14) Chiu, P.; Li, S. Org. Lett. 2004, 6, 613-616.
(15) Proton H3′, which is cis to the 2-alkyl chain on the heterocycle,
consistently resonates at about 2.1 ppm (as a doublet of doublets) in
all Nazarov compounds. NOE cross-peaks between the isopropyl
protons and H3′ are diagnostic of the cis relative stereochemistry. In
the 1H NMR of 28 the signals of two rotamers in a 3:1 ratio are also
present, as demonstrated by recording 1H NMR spectra at different
temperatures.
(16) This corresponds to an “interrupted” Nazarov reaction. See, for
example: (a) Bender, J. A.; Blize, A. E.; Browder, C. C.; Giese, S.; West,
F. G. J. Org. Chem. 1998, 63, 2430-2431. (b) Bender, J. A.; Arif, A.
M.; West, F. G. J. Am. Chem. Soc. 1999, 121, 7443-7444.
(17) Luker, T.; Hiemstra, H.; Speckamp, W. N. J. Org. Chem. 1997,
62, 8131-8140.
4544 J. Org. Chem., Vol. 70, No. 11, 2005

intermediate in the synthesis of roseophilin was obtained
in eight step from known compound 9 (3% overall yield).
In conclusion, in this work we have experimented with
and compared three different procedures to obtain prop-
erly substituted divinyl ketones in which one of the
double bonds is embedded in a five-membered heterocy-
clic structure, suitable to produce cyclopenta-fused pyr-
role derivatives by the acid-catalyzed Nazarov reaction.
Despite the inherent difficulties in working with pyrro-
lidinone-derived vinyl triflates, in one case a convergent
procedure based on the umpolung to a vinylstannane
gave acceptable yields of the requisite divinyl ketone. The
successive Nazarov reaction of these compounds on
treatment with pure TFA afforded 2,4-cis-disubstituted
2,3,4,5-tetrahydro-1H-cyclopenta[b]pyrrolones with high
stereocontrol. One of these was oxidized to the corre-
sponding 4,5-dihydro-1H-cyclopenta[b]pyrrol-6-one, thus
obtaining an enantiopure key intermediate in the total
synthesis of roseophilin.
Acknowledgment. We thank MIUR and University
of Florence for financial support. Mr. Maurizio Passa-
ponti and Mrs. Brunella Innocenti are acknowledged for
their technical support. Prof. P. L. Fuchs is acknowl-
edged for providing the conditions for the N-iodosuc-
cinimide oxidation. Ente Cassa di Risparmio di Firenze
is acknowledged for granting a 400 MHz NMR spec-
trometer.
Supporting Information Available: Full experimental
details and compound characterization; copies of the 1H NMR
spectra of compounds 27, 28, 10-12, 21, 24, 7, 6, 5, and 14.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(18) Shim, Y. K.; Youn, J. I.; Chun, J. S.; Park, T. H.; Kim, M. H.;
Kim, W. J. Synthesis 1990, 753-754. Oxidation by NIS in CCl4 (ref
2a) provided 5 in lower yield.
(19) The 13C NMR spectrum of (S)-5 is identical to that of the
racemic material kindly provided by Prof. Fuchs.
JO0504058
J. Org. Chem, Vol. 70, No. 11, 2005 4545